Ever since my (not-rooted stock Android) Nokia 6.1 phone was upgraded to Android 10, last January, I have a performance issue with it: A couple of times per day it becomes near or completely unusable, often permanently until a hard restart. Also this causes high battery usage.

However, using a battery analysis app didn't turn up any specific suspect.

Recently I discovered adb shell, and when the phone starts behaving badly,

  • top showed me that the CPU-hogging process at such moments is kswapd:

    Tasks: 681 total,   1 running, 680 sleeping,   0 stopped,   0 zombie
      Mem:      2.6G total,      2.4G used,      198M free,       11M buffers
     Swap:      1.0G total,      1.0G used,      112K free,      311M cached
    800%cpu  38%user  10%nice  52%sys 684%idle  10%iow   2%irq   4%sirq   0%host
      PID USER         PR  NI VIRT  RES  SHR S[%CPU] %MEM     TIME+ ARGS
     1739 system       18  -2 7.9G 144M  47M S 12.3   5.2  13:37.24 system_server
      148 root         20   0    0    0    0 S  8.6   0.0   3:50.15 [kswapd0]
      564 logd         30  10 168M  10M 0.9M S  7.3   0.3   1:11.69 logd
    28119 shell        20   0  35M 3.1M 1.7M R  5.0   0.1   0:00.45 top
      705 root         20   0 5.1G  44M  24M S  5.0   1.6   0:54.88 zygote64
  • top -s 6 (sorts by memory use, RES) showed me that the main memory-using process at such moments is com.android.vending (= Google Play Store).

    Tasks: 690 total,   1 running, 689 sleeping,   0 stopped,   0 zombie
      Mem:      2.6G total,      2.4G used,      180M free,       17M buffers
     Swap:      1.0G total,      0.9G used,      2.1M free,      412M cached
    800%cpu   9%user   3%nice  13%sys 772%idle   1%iow   2%irq   1%sirq   0%host
      PID USER         PR  NI VIRT [RES] SHR S %CPU  %MEM     TIME+ ARGS
     5296 u0_a105      20   0 2.1G 472M  24M S  0.0  17.2   1:18.80 com.android.vending
     1739 system       18  -2 7.9G 150M  53M S  3.0   5.4  13:12.22 system_server
     2233 u0_a111      20   0 6.1G 108M  56M S  0.6   3.9   2:03.84 com.android.systemui
     7120 u0_a161      20   0 7.7G  93M  45M S  0.6   3.4   8:22.83 com.google.android.youtube
     2931 u0_a90       20   0 5.9G  83M  36M S  0.0   3.0   3:05.21 com.google.android.gms.persistent
    26244 u0_a90       20   0 5.2G  74M  39M S  0.0   2.7   0:02.38 com.google.android.gms.ui

[Side note. That seems to explain why battery analysis didn't help: Presumably kswapd CPU usage is ascribed to general kernel processing, not to this specific app.]

And that app is really the cause: Stopping that Google Play Store process (e.g., using am force-stop com.android.vending) quickly brings performance back to normal levels.

As a workaround, I tried to disable automatic updates in Google Play Store, and that seems to have improved things a little, but the issue is definitely still there.

So now, as another workaround, a couple of times per day I hard-stop that application, which helps to prevent it usually.

My vague question is: How to I get further with this issue?

  • Is this a known issue that has already been reported somewhere? Is the cause known? (E.g., is this triggered by the Google Play Store app detecting updates?)
  • Can I report it to Nokia in some bug tracker? Can I report it to Google somewhere?
  • What kind of additional information/dumps/logs would help to investigate this further?
  • Are there any other/better workarounds? (Disable some feature, go back to an older version of Google Play Store, clear some cache or force something to be regenerated, disable swap, whatever...)
  • Is there some way to automate the workaround I currently have? (I'm already using Automate for some other things, but any automated solution apparently requires my device to be rooted.)

(Note that my device is not rooted, and I would like to try to keep it that way. But I will consider rooting it, if there some kind of solution or workaround that requires that.

If it helps, phone model is TA-1043, currently running build number 00WW_4_10C_SP04.)

  • 118 views over 4 months and no suggestion for even one of my points? Anything would help me here. (It doesn't seem helpful to go back to factory settings, since it's install the same apps and run into the same issue?) Sep 22, 2020 at 17:11
  • It may not be germane, or even helpful, but I started seeing this on my Nexus 4 yesterday and a search brought this up.
    – pete
    Sep 23, 2020 at 16:49
  • For the record, this specific issue is moot for me, because within a few months of asking this question, this device stopped working with an apparently unrelated hardware issue. (Even after reset to factory defaults, it would spontaneously restart once a day or even more often.) I still would appreciate 'where to report' and 'how to debug' pointers and advice! For example, because I again bought a similar phone, and I'm now debugging another issue 🙂; see android.stackexchange.com/q/249383/37933. Dec 15, 2022 at 7:54


You must log in to answer this question.

Browse other questions tagged .